Coronary heart disease is a major cause of death, particularly in western world countries where the populations are generally well-nourished. Many factors are implicated as risk factors for this disease including obesity, smoking, genetic predisposition, diet, hypertension, and cholesterol.
Dairy products, especially milk, are a major contributor to the dietary intake of humans, again particularly in western world populations. Milk contains numerous components of nutritional and health benefit. Calcium is one example. However, milk is also a significant source of dietary fat. It is widely accepted that saturated fats found in milk are a risk factor for coronary heart disease. However, an additional risk factor present in some bovine milk unrelated to the fat content has been discovered. What is entirely surprising is the source of the risk. The source is not dependent on the fat content of milk. Instead, it is a milk protein, β-casein, which is linked to coronary heart disease.
A number of variants of milk proteins have been identified. Initially, three variants of β-casein were discovered (Aschaffenburg, 1961) and were denoted A, B and C. It was later found that the A variant could be resolved into A1, A2 and A3 by gel electrophoresis (Peterson et. al. 1966). The β-casein variants now known are A1, A2, A3, B, C, D, E and F, with A1 and A2 being present in milk in the highest proportions. It is anticipated that other variants may be identified in the future.
The milk protein β-casein A1 has been determined to represent the risk factor in bovine milk that is linked to coronary heart disease, or at least is the principal risk factor. This determination forms the basis of the present invention.
There is no relationship between the fat content of milk and β-casein genotype in cows. Therefore, selecting cattle on the basis of milk fat content will not identify which bovines produce the novel risk factor, namely the specific β-casein variant, in their milk.
There is no significant difference in the fat content of milk produced by cows which are homozygous for the β-casein A1 allele (i.e. A1A1) and cows which are homozygous for the β-casein A2 allele (i.e. A2A2). This is apparent from studies reported in the literature.
Ng-Kwai-Hang has carried out several studies. One study (Ng-Kwai-Hang et. al., 1990) suggested that milk containing β-casein A1 rather than β-casein A2 may have a slightly higher fat content. However, these differences were very small. The differences between milk from A1 homozygous cows and milk from A2 homozygous cows were 0.05% (for the first lactation period), 0.07% (for the second lactation period), and 0.04% (for the third lactation period).
In another study (in 1995), Ng-Kwai-Hang (in an abstract cited by Jakob et. al., 1997) found the opposite effect (i.e. the A1A1 product had a lower fat content than the A2A2 product). Thus, the 1995 Ng-Kwai-Hang study directly contradicts the Ng-Kwai-Hang, et. al., 1990 study.
McLean et. al., 1984 (McLean) also reported that there was no significant difference in the fat content of milk from cows of A1A1 and A2A2 genotypes (mean±standard error: 45.8±2.6 g/l for milk of A1A1 cows and 48.6±1.9 g/l for milk of A2A2 cows).
Aleandri et al., 1990 (Aleandri), shows in Table 5 that the least squares estimates of the effects of different genotypes and their standard errors on fat percentage in milk are 0.12±0.09 for A1A1 cows and 0.07±0.09 for A2A2 cows. Taking into account the standard error for the test, Aleandri indicates that the effects of A1A1 and A2A2 genotypes on milk fat content are equivalent.
Bovenhuis et. al., 1992 (Bovenhuis), highlights statistical problems associated with the way in which the genotype effects on fat percentages in milk are studied and documented. It is stated that the ordinary least squares estimates may be biased. Bovenhuis points out that the analysis of the effect of a particular genotype on various characteristics of milk is complex in nature and may, among other things, be affected by other genes which may be linked to the gene under study. Bovenhuis attempts to take into account the above variables and to overcome statistical problems by using an animal model method.
Table 3 of Bovenhuis indicates that, for a statistical model in which each milk protein gene is analysed separately and the A1A1 cows designated as being the standard (i.e. given a value of 0% fat attributable to the genotype), the A2A2 genotype was estimated not to contribute (i.e. 0%) to the fat content of the milk of the animals harbouring that genotype when compared to the A1A1 genotype. The standard error of the test is recorded as 0.02%. Where a statistical model was used in which all milk protein genes were analysed simultaneously (Table 4 of Bovenhuis) and the A1A1 genotype was again designated as being the standard (at 0% fat content attributable to the A1A1 genotype), the A2A2 genotype was estimated to contribute to the fat content of the milk at −0.01% when compared with the A1A1 genotype. In this study a standard error of 0.02 was designated. Taking into account the standard error of the tests these results indicate that the A2A2 genotype contributes to the fat content of milk in an equivalent manner to the genotype A1A1. Gonyon et. al., 1987 (Gonyon) reached the same conclusion as Bovenhuis.
The level of individual components in milk is influenced by both the genotype and the environment. That is, the variation between animals in milk output or milk composition is due to both genotypic and phenotypic factors. For example, Bassette et. al., 1988 (Bassette) indicates that the composition of bovine milk may be influenced by a number of environmental factors and conditions other than genetic factors. Environmental factors may impact on milk production and the constituents contained within the milk (including fat content). For example, changes in milk composition occur due to:                the stage of lactation (e.g., the fat content of colostrum is often higher; the concentration of fat changes over a period of many weeks as the cow goes through lactation);        the age of the cow and the number of previous lactations;        the nutrition of the cow including the type and composition of feed consumed by the cow;        seasonal variations;        the environmental temperature at which the cows are held;        variations due to the milking procedure (e.g., the fat content of milk tends to increase during the milking process which means that for an incomplete milking the fat content would generally be lower than normal and for a complete milking the fat content will be higher than normal); and        milking at different times of the day.        
It is therefore apparent from the studies in this field that a person skilled in the relevant area of technology would not find a link between the fat content of milk and the β-casein genotype of the milk-producing bovines from which that milk is produced. The milk protein β-casein A1 has now been identified as a risk factor linked to coronary heart disease in its own right.
Epidemiological evidence strongly suggests that dietary β-casein A1 is harmful to human health. For example, WO 96/14577 describes the impact of β-casein A1 on Type I diabetes. The epidemiological evidence suggests that β-casein A1 stimulates diabetogenic activity in humans. Furthermore, WO 96/14577 describes the induction of autoimmune, or Type I, diabetes in the non-obese diabetic (NOD) mouse model by way of consumption of β-casein A1. The invention described focuses on reducing the risk of contracting Type I diabetes in a susceptible individual by restricting the milk or milk product intake of that individual to milk containing only non-diabetogenic β-casein variants.
Beales et al., 2002 (Beales) describes an investigation of whether β-casein A1 is more diabetogenic than β-casein A2. The results were not conclusive, although β-casein A1 was found to be more diabetogenic than β-casein A2 in certain rats (BB rats) fed a soy isolate based infant formula (ProSobee).
WO 96/36239 is the PCT publication for PCT/NZ96/00029 from which U.S. Ser. No. 09/906,614 and this application are derived. WO 96/36239 advocates the avoidance of β-casein A1 in the human diet. Epidemiological evidence establishes a correlation between the consumption of β-casein A1 in various populations and the incidence of coronary heart disease. In particular, the invention of U.S. Ser. No. 09/906,614 focuses on the avoidance of β-casein A1 by selecting milking cows by testing genetic material for DNA encoding the various β-caseins.
McLachlan, 2001 (McLachlan) presents data reporting the link between the consumption of β-casein A1 and both type 1 diabetes and coronary heart disease.
Thus, the link between β-casein A1 and diabetes and the link between β-casein A1 and coronary heart disease have both been well documented. Additionally, it has recently been shown that the deleterious effects of β-casein A1 now extend to neurological disorders. WO 02/19832 describes the avoidance of inducing or aggravating a neurological or mental disorder by providing milk which does not contain any “histidine variant”. A histidine variant is defined in that document as a β-casein variant which has histidine at position 67. β-Casein A1 is such a histidine variant.
WO 01/00047 discloses a dietary supplement comprising a milk product which contains predominantly the A2 variant of the β-caseins, and which is fortified with a compound (or compounds) that lowers human plasma homocyst(e)ine levels. Such compounds may include betaine, cobalamin, folic acid or pyridoxine. Two possible advantages to this are described. Firstly, the replacement of β-casein A1 in the supplement with β-casein A2 will lower the risk of diabetes. Secondly, it is desirable to lower homocyst(e)ine levels, since these are highly correlated with coronary heart disease, and homocyst(e)ine is also a recognised risk factor for atherosclerosis. Thus, there is a combined approach, using the homocyst(e)ine strategy together with the diabetes strategy, for controlling vascular disease.
Laugesen and Elliot, 2003 (Laugesen) present further evidence of the link between the consumption of β-casein A1 and both coronary heart disease and diabetes. In this epidemiology study, Laugesen concludes that the consumption per capita of β-casein A1 is significantly and positively correlated with ischaemic heart disease in 20 affluent countries. The study also confirms earlier findings made by Elliot that the consumption per capita of β-casein A1 is correlated with type 1 diabetes mellitus.
A method of producing milk substantially free of β-casein A1 is described and claimed in the inventor's U.S. patent application Ser. No. 09/906,807. The invention of that application relates to the testing of DNA or RNA from cells obtained from lactating bovines. Those bovines which do not have any DNA or RNA encoding for β-casein A1 are then selected and milked.
The subject matter of this application relates to the breeding of bovine bulls that do not have DNA encoding for β-casein A1 with bovine cows that do not have DNA encoding for β-casein A1. The progeny cows therefore do not produce β-casein A1 in their milk. These cows are then milked to provide milk suitable for use in the treatment or prevention of coronary heart disease.
It is therefore an object of this invention to provide a method of producing milk substantially free of β-casein A1 suitable for use in the prevention or treatment of coronary heart disease, or to at least provide a useful alternative.